Add custom_objects in the function layer_test #36273
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
layer test for custom layer
Usage:
```
testing_utils.layer_test(
SyncBatchNormalization,
# keras.layers.BatchNormalization,
kwargs={
'momentum': 0.9,
'epsilon': 0.1,
'gamma_regularizer': keras.regularizers.l2(0.01),
'beta_regularizer': keras.regularizers.l2(0.01)
},
input_shape=(3, 4, 2),
custom_objects={'SyncBatchNormalization': SyncBatchNormalization})
```
|
Thanks for your pull request. It looks like this may be your first contribution to a Google open source project (if not, look below for help). Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA). 📝 Please visit https://cla.developers.google.com/ to sign. Once you've signed (or fixed any issues), please reply here with What to do if you already signed the CLAIndividual signers
Corporate signers
ℹ️ Googlers: Go here for more info. |
|
@googlebot I signed it! |
|
CLAs look good, thanks! ℹ️ Googlers: Go here for more info. |
|
Hi Mokke, Just realized I accidentally commented on the issue instead of this PR: layer_test is a totally internal method, we don't want external users relying on it. This change seems explicitly designed for external users of Keras implementing their own layers and trying to use layer_test. The layer_test utility isn't an effective test of layer behavior, it just makes sure certain things don't crash. You would be better served writing your own set of unit tests that actually verify layer behavior (and that make sure training, evaluation, etc. work correctly). That said, I'll add @yhliang2018 to see if she thinks that there are some sort of testing utilities we should be exposing to external users in Keras' API. |
|
@MokkeMeguru Can you please check tomerk's comments and keep us posted? Thanks! |
|
I waited @yhliang2018 's comment. |
yhliang2018
reviewed
Feb 27, 2020
| @@ -93,7 +93,9 @@ def layer_test(layer_cls, kwargs=None, input_shape=None, input_dtype=None, | |||
| string or integer values. | |||
| adapt_data: Optional data for an 'adapt' call. If None, adapt() will not | |||
| be tested for this layer. This is only relevant for PreprocessingLayers. | |||
|
|
|||
| custom_objects: Optional Objects for your custom layer. If you write a | |||
| your layer, you can use this variable. | |||
There was a problem hiding this comment.
nit: custom_objects: Optional dictionary mapping name strings to custom objects in the layer class. This is helpful to test a custom layer.
There was a problem hiding this comment.
Thanks your help!
There was a problem hiding this comment.
I think @yhliang2018 meant that you should update this description to instead say:
custom_objects: Optional dictionary mapping name strings to custom objects in the layer class. This is helpful for testing custom layers.
Then we can go ahead and approve this.
|
Thanks for the PR! I think it's okay to add an extra arg to |
tensorflowbutler
removed
the
stat:awaiting response
|
response? Should I tell you anything more? |
|
@MokkeMeguru Can you please address the reviewers comments and resolve conflicts? Thanks! |
|
I fixed import error for organizing imports. |
yhliang2018
previously approved these changes
Mar 6, 2020
tensorflow-bot
bot
added
kokoro:force-run
|
I don't know why this pr is not accepted by Unity Sanity with pylint error. logs. |
|
this means we should fix the code, to ? (two space to four space) |
TensorFlow's linting requires four spaces.
|
I updated PR for passing your lint test. |
tomerk
approved these changes
Mar 9, 2020
tensorflow-bot
bot
added
kokoro:force-run
|
I don't know why windows build will be crash. |
gbaned
added
kokoro:force-run
|
I updated the description because I think the curly braces in the code snippets are causing issues with the merging. @gbaned how do we restart the merging? Usage: |
|
Thanks a lot! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add a custom_objects argument to layer_test, to enable running layer_test for custom layers.
reference: #36272